The use of ChatGPT in service recovery: Compensating customers

Abstract

Determining the appropriate compensation for customers is a crucial decision, as it may result in the wasting of resources and further exacerbating customer frustration. Making the right compensation decision requires a great deal of knowledge and expertise about the customers and their service encounters, as well as taking both the customers' and the service business's interests into account. This study investigates the usability of ChatGPT, as a generative AI tool, in identifying the severity of service failures for customers and producing an effective and efficient compensation suggestion accordingly. The two surveys in the study, carried out in two stages with 298 hotel customers and 54 managers from 5-star hotels, established that no single compensation strategy developed by ChatGPT can satisfy most of the customers, and a combination of compensation strategies needs to be used. The study has important theoretical and practical implications both regarding the field of generative AI, in terms of developing business solutions, and for the service recovery and compensation literature. © 2025 Elsevier B.V., All rights reserved.